Diagnostic yield of endoscopic ultrasonography in patients with intermediate or high likelihood of choledocholithiasis: a retrospective study from one university-based endoscopy center

BACKGROUND: Diagnosis of choledocholithiasis requires clinical manifestations and imaging examination findings suggesting a stone in the common bile duct (CBD), but these factors are not highly sensitive or specific. The choledocholithiasis management algorithm proposed by the American Society for Gastrointestinal Endoscopy (ASGE) may not be appropriate for patients who fulfill the clinical criteria for a high likelihood of choledocholithiasis. Endoscopic ultrasonography (EUS) may replace endoscopic retrograde cholangiopancreatography (ERCP) for the detection of CBD stones in all patients. The aims of this study were to determine the diagnostic yield and optimal timing of EUS in patients with an intermediate or high likelihood of choledocholithiasis requiring therapeutic ERCP. METHODS: Patients with suspected choledocholithiasis who underwent EUS between June 2009 and January 2012 were retrospectively reviewed. The patients were divided into two groups based on the likelihood of choledocholithiasis according to the clinical predictors described by the ASGE guidelines: an intermediate likelihood group and a high likelihood group. The demographic data, clinical manifestations at presentation, blood test results, EUS and ERCP findings, and clinical manifestations during the follow-up period were recorded and analyzed. RESULTS: Ninety-three patients were enrolled in the study (52.7% in the intermediate likelihood group and 47.3% in the high likelihood group). CBD stones were detected in 22.44% of patients in the intermediate likelihood group and 38.63% of patients in the high likelihood group. EUS had a sensitivity of 100% and specificity of 80% for detection of CBD stones. An alkaline phosphatase level of >133 mg/dL (area under the curve, 0.576) was the only factor that was significantly associated with detection of CBD stones in patients who underwent EUS >7 days after the initial clinical presentation (odds ratio 4.87, p = 0.01). CONCLUSIONS: EUS is an accurate diagnostic tool for the detection of CBD stones, and can prevent the unnecessary use of ERCP. This study found that use of clinical criteria alone might not provide a good prediction of the presence of CBD stones, even in patients who fulfill the criteria for a high likelihood of choledocholithiasis.